AskAway

Inspiration

Isn’t it frustrating to go through entire book to just grasp a single specific topic? So we decided to choose NLP model that can be trained on specific pdf and can find you that exact information you need.

What it does

Ask-Away is essentially a large language model that can undergo training on a local machine without an actual internet connection. It provides a user-friendly AI interface, allowing users to train the model on a particular book and then pose questions related to that book.

How we built it

The project uses the LangChain library to handle data ingestion, model training loops and managing the model lifecycle. The ChainLit framework is used to package and deploy the LLM inference server for easy access through a web UI or APIs.

Challenges we ran into

Environment Setup Deployment Library Managements; LangChain problems

Accomplishments that we're proud of

We accomplished what we started with. We wanted to design a book summarization feature using chatbots from the input pdf we wanted

What we learned

We learned different libraries in python like pypdf langchain,torch,accelerate,bitsandbytes,transformers,sentence_transformers,faiss_cpu,chainlit

What's next for AskAway

We are looking to deploy a web application that enables users to upload PDF book files via a user-friendly web interface. Upon uploading, the application's integrated chatbot will promptly generate concise summaries based on the uploaded content's category. This streamlined process ensures efficient access to summarized information, enhancing user experience and knowledge retrieval. The development of this solution adheres to a time-sensitive framework, emphasizing the importance of timely and efficient delivery.

Built With

• chainlit
• langchain
• python